system_message = "You are a text classifier and are only allowed to respond with 1, 2, or 0."

sentiment_message = """You are a text classifier that determines whether the sentiment of a tweet is positive, negative, or neutral.

For the text sample below, return one of the following labels:
1 if the tweet's sentiment is positive.
2 if the tweet's sentiment is negative.
0 if the tweet's sentiment is neutral.

Here is the tweet:\n{text}

Do not explain your reasoning. Respond only with the number representing the correct label.
"""

valence_message = """You are a text classifier that determines whether the emotions expressed in a tweet are primarily positive, negative, or neutral.

For the tweet below, return one of the following labels:
1 if the tweet expresses primarily positive emotions.
2 if the tweet expresses primarily negative emotions.
0 if the tweet is neutral or does not express emotions.

Here is the tweet:\n{text}

Do not explain your reasoning. Respond only with the number representing the correct label.
"""

stance_message = """You are a text classifier that determines whether the author of a tweet supports, opposes, or is neutral towards {target}.

For each tweet, return one of the following labels:
1 if the author most likely supports {target}.
2 if the author most likely opposes {target}.
0 if the tweet does not mention {target}, or if there is no clear indication of the author's stance toward {target}.

Here is the tweet:\n{text}

Do not explain your reasoning. Respond only with the number representing the correct label.
"""

#sentiment_message = """Is the sentiment of this text positive, neutral, or negative? Answer only with a number: 1 if positive, 2 if negative, and 0 if neutral. 
#Here is the text: 
#{text}"""